+ (,1 1/,1( 2
Citation: 153 U. Pa. L. Rev. 1883 2004-2005 Content downloaded/printed from
HeinOnline (http://heinonline.org) Mon Jun 25 09:47:47 2012
-- Your use of this HeinOnline PDF indicates your acceptance of HeinOnline's Terms and Conditions of the license
agreement available at http://heinonline.org/HOL/License -- The search text of this PDF is generated from
uncorrected OCR text.
-- To obtain permission to use this article beyond the scope of your HeinOnline license, please use:
https://www.copyright.com/ccc/basicSearch.do?
&operation=go&searchType=0
&lastSearch=simple&all=on&titleOrStdNo=0041-9907
Retrieved from DiscoverArchive, Vanderbilt University’s Institutional Repository
This work was originally published in 153 U. Pa. L. Rev. 1883 2004-2005
CHOICE OF LAW FOR INTERNET TRANSACTIONS: THE UNEASY CASE FOR ONLINE CONSUMER PROTECTION
ERIN ANN O'HARAt INTRODUCTION
An impressively large number of consumer transactions are occur- ring online these days. Today, millions of consumers buy billions of dollars worth of goods online in a single year, and the numbers con- tinue to grow. Presumably, the benefits to online purchases are all about efficiency. Vendors can conserve on the costs of maintaining stores and hiring employees to properly staff them. Consumers can conserve on the time and travel costs associated with shopping; and, unlike their offline counterparts, the online stores never close. To this extent, Internet transactions have been a huge success.
Despite this success, states might not be doing all that they can to increase vendor competition and thereby decrease the prices paid by consumers. Empirical studies indicate that significant price disparities for the same or similar products still exist online. Moreover, unknown and new online vendors typically must pay an extra five to ten percent of the consumer's purchase price to third-party intermediaries who help ensure that the transaction goes smoothly. The numbers suggest that known vendors might well take in more revenues at lower ex- pense to provide the same goods to consumers that unknown vendors provide.
This Essay explores the possibility that the market for online pur- chases fails to work as efficiently as it can because consumers lack trust in unknown vendors, and it argues that consumer distrust in unknown vendors can and often does take the form of categorical avoidance of other unknown vendors. This avoidance of unknown vendors as a class results from the fact that trust and distrust, as cognitive phenom-
t Professor of Law, Vanderbilt University School of Law. I received generous re- search support for this project from Vanderbilt University School of Law. Erica Beecher-Monas, Allison Danner, Tracey George, John Goldberg, David Gordon, Chris Guthrie, Larry Helfer, Steve Hetcher, Owen Jones, Kevin McCabe, Greg Mitchell, Eli- nor Ostrom, Robert Rasmussen, Alan Stein, Peter Swire, Mike Vandenbergh, Paul Zak, and especially Claire Hill have provided very helpful comments and suggestions for this paper. Thank you also toJulie Reed for valuable research assistance.
(1883)
1884 UNIVERSITY OFPENNSYLVANIA LAWREVIEW [Vol. 153:1883 ena, are subject to the same biases and limitations as are other cogni- tive phenomena. Unknown vendors often are willing to incur some costs to signal their trustworthiness to individual consumers. Unless the other unknown vendors adopt similar trust-promoting practices, however, the unknown vendor who incurs these costs might not expe- rience a proportionate increase in consumer trust. Instead, the invest- ing vendor can continue to be plagued by consumers' categorical re- sistance to unknown vendors. Put differently, consumer distrust of unknown vendors represents a negative externality that the actions of a single vendor are often powerless to resolve.
To be sure, some unknown vendors have been successful in adopt- ing online sales practices that promote consumer trust. When suc- cessful, previously unknown vendors become known vendors for pur- poses of their relevant consumer market. For many goods and consumer market niches, however, unknown vendors remain at a competitive disadvantage relative to known vendors, and the question explored in this Essay is what, if anything, contract law should do to enable unknown vendors to compete more effectively. More specifi- cally, this Essay explores the extent to which the contract doctrine ap- plied to online transactions, if carefully developed and uniformly ap- plied by courts, could serve a coordinating function for vendor behavior that might work to promote consumer trust in unknown vendors. If contract doctrine could successfully aid trust enhance- ment, then unknown vendors could compete more effectively online at lower expense, and consumer welfare would increase. For reasons explained in this Essay, contract doctrine may prove to be too indirect a tool for promoting consumer trust, but some experimentation could nevertheless be worthwhile.
Part I of this Essay describes trust as a cognitive phenomenon and presents an informal model of the relationship between trust and the law. Part II describes the problem of online trust by consumer pur- chasers. Part III explores the possible ways that contract doctrines can work to enhance consumer trust. It explores conspicuousness and clarity rules designed to better notify consumers of the terms of stan- dard-form contracts, advocates close scrutiny of arbitration clauses and blanket warranty exclusions, promotes generous damage awards, and explores several possible rules regarding the provision of infor- mation and assent to terms on websites. Each of the rules or doctrines explored exists in at least one jurisdiction, but this Essay recommends that all of the proposed legal treatments be adopted in all jurisdic- tions. In the United States, this bundle of contract protections will
2005] CHOICE OF LA W FOR INTERNET TRANSA CTIONS
likely be more generous to consumers than the bundle that currently exists in any one jurisdiction, but the enhanced protections might help facilitate online competition and therefore be warranted.
Part IV explores ways that these contract rules can be applied uni- formly across jurisdictions to better promote consumer trust and briefly examines the success of harmonization efforts. Part IV also ad- vocates that in the interim states should follow a place-of-consumer- residence or place-of-delivery rule for choice of law in order to pro- vide the best incentives for states to adopt efficient protections for on- line consumer contracts. Even if harmonization of the online con- sumer protections could be achieved, however, it is not clear that consumer trust in online vendors would be significantly enhanced.
To avoid the costs of consumer opportunism, the reforms advocated in this Essay are quite modest, and perhaps too modest, given the re- moteness of contract law to consumer decisions, to have the effect of enhancing consumer protection in unknown vendors. Nevertheless, a summary section in Part V argues that experimentation with these modest reforms could prove worthwhile.
I. TRUST AND ITS RELATIONSHIP TO LAW
A. Trust as a Cognitive Phenomenon
Law and psychology is an exciting field these days. Legal scholars have focused recently on empirical and theoretical questions sur- rounding the incorporation of insights about the psychology of hu- man actors.' The motivation for much of the early work in one sub-
] See generally BEHAVIORAL LAW AND ECONOMICS (Cass R. Sunstein ed., 2000);
LAWRENCE S. WRIGHTSMAN ET AL., PSYCHOLOGY & THE LEGAL SYSTEM (4th ed. 1998);
Lynn A- Baker & Robert E. Emery, When Every Relationship Is Above Average: Perceptions and Expectations of Divorce at the Time of Marriage, 17 LAW & HUM. BEHAv. 439 (1993);
Chris Guthrie, Better Settle than Sony: The Regret Aversion Theory of Litigation Behavior, 1999 U. ILL. L. REV. 43; Chris Guthrie, Framing Frivolous Litigation: A Psychological The- ory, 67 U. CHI. L. REV. 163 (2000); Chris Guthrie et al., Judging by Heuristic: Cognitive Illusions in Judicial Decision Making, 86 JUDICATURE 44 (2002); Chris Guthrie, Prospect Theory, Risk Preference, and the Law, 97 Nw. U. L. REV. 1115 (2003); Russell Korobkin, The Efficiency of Managed Care "Patient Protection" Laws: Incomplete Contracts, Bounded Ra- tionality, and Market Failure, 85 CORNELL L. REV. 1 (1999); Russell Korobkin, The En- dowment Effect and Legal Analysis, 97 Nw. U. L. REV. 1227 (2003); Russell Korobkin &
Chris Guthrie, Psychological Barriers to Litigation Settlement: An Experimental Approach, 93 MICH. L. REV. 107 (1994); Gregory Mitchell, Taking Behavioralism Too Seriously? The
Unwarranted Pessimism of the New Behavioral Analysis of Law, 43 WM. & MARY L. REV. 1907 (2002); Janice Nadler, No Need to Shout: Bus Sweeps and the Psychology of Coercion, 2002 SUP. CT. REV. 153; Janice Nadler & Mary R. Rose, Victim Impact Testimony and the Psy-
1885
1886 UNIVERSITY OFPENNSYLVANIA LAWREVIEW [Vol. 153:1883 field of law and psychology, behavioral decision theory, was to chal- lenge the rational actor assumption underlying virtually all of the scholarship in law and economics. More recently, these alternative theories of human decision making have become fruitful areas of study in their own right, and have generated a lively debate about the policy implications of this new knowledge.3
Separately, a rich body of norms scholarship has recently devel-S 4
oped that attempts to explain the extralegal influences on behavior.
According to this scholarship, norms come in two forms: social norms and internal (or personal) norms.5 Social norms are premised on the idea that humans evolved in groups as a necessary means for survival 6
chology of Punishment, 88 CORNELL L. REV. 419 (2003);JeffreyJ. Rachlinski & Cynthia R.
Farina, Cognitive Psychology and Optimal Government Design, 87 CORNELL L. REV. 549 (2002); Jeffrey J. Rachlinski & Forest Jourden, The Cognitive Components of Punishment, 88 CORNELL L. REV. 457 (2003).
2 See generallyJames A. Fanto, Quasi-Rationality in Action: A Study of Psychological Fac- tors in Merger Decision-Making, 62 OHIO ST. L.J. 1333 (2001); Chris Guthrie et al., Inside the Judicial Mind, 86 CORNELL L. REV. 777 (2001); ChristineJolls et al., A Behavioral Ap- proach to Law and Economics, 50 STAN. L. REV. 1471 (1998); Russell B. Korobkin &
Thomas S. Ulen, Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics, 88 CAL. L. REV. 1051 (2000); Jeffrey J. Rachlinski, The "New" Law and Psychology: A Reply to Critics, Skeptics, and Cautious Supporters, 85 CORNELL L. REV.
739 (2000).
3 Compare Jeffrey J. Rachlinski, The Uncertain Psychological Case for Paternalism, 97 Nw. U. L. REV. 1165, 1177-1206 (2003) (discussing the notion that behavioral decision theory implicates legal constraints on individual choice), with Colin Camerer et al., Regulation for Conservatives: Behavioral Economics and the Case for "Asymmetric Paternalism,"
151 U. PA. L. REV. 1211 (2003) (advocating regulation that interferes only with choices of those whose decisions are tainted by irrationality), and Cass R. Sunstein & Richard H. Thaler, Libertarian Paternalism Is Not an Oxymoron, 70 U. CHI. L. REV. 1159 (2003) (proposing a series of default rules that steer people away from irrational choices and encouraging government to frame sets of choices in a manner that encourages better decision making), and Gregory Mitchell, Libertarian Paternalism Is an Oxymoron, 99 Nw.
U. L. REV. (forthcoming 2005) (advocating procedural mechanisms that increase probablitiy that people will make choices without framing and other irrational influ- ences).
4 See generally ROBERT C. ELLICKSON, ORDER WITHOUT LAW: How NEIGHBORS SETrLE DISPUTES (1991); ERIC A. POSNER, LAW AND SOCIAL NORMS (2000); Richard H.
McAdams, The Origin, Development, and Regulation of Norms, 96 MICH. L. REV. 338 (1997); Cass R. Sunstein, Social Norms and Social Roles, 96 COLUM. L. REV. 903 (1996).
5 See, e.g., Michael P. Vandenbergh, From Smokestack to SUV: The Individual as Regu- lated Entity in the New Era of Environmental Law, 57 VAND. L. REV. 515, 596 (2004) (defin- ing the two types of norms).
6 See, e.g., MATT RIDLEY, THE ORIGINS OF VIRTUE: HUMAN INSTINCTS AND THE EVOLUTION OF COOPERATION 69 (1996) (stating that hunter-gatherers tended to band together in groups of about 150 persons); PAUL H. RUBIN, DARWINIAN POLITICS: THE EVOLUTIONARY ORIGIN OF FREEDOM 7 (2002) ("Based on our brain size compared with other primates, it appears that we are adapted to groups of 150 ....").
2005] CHOICE OF LA WFOR INTERNET TRANSACTIONS
and, as a result, they are often highly influenced by the reactions of other group members to their behavior and views. No doubt we seek both formal and informal membership in those groups that are ac- cepting of our views and our way of life. But the group itself also in- fluences, at least at the margins, the views and actions of its members.8 Any legal efforts to control, influence, or modify behavior must take these social norms and private reward and sanction systems into ac- count. Norms scholars have also focused on internal (or personal) norms, or rules of conduct that individuals develop as part of their moral and ethical beliefs about appropriate conduct.9 The presence or absence of internal norms can also have profound effects on the degree of legal sanction necessary to effectively influence behavior.
Legal rules and sanctions can in turn affect the development and strength of both internal and social norms, because the positive law provides a signal about the harmfulness or desirability of behavior and about the normative judgment of that behavior by the larger so- cial/political/economic group to which a member belongs.10
As these two bodies of scholarship-norms theory and behavioral decision theory-have developed, a few legal scholars have turned their attention to interpersonal trust, a subject that incorporates in- sights from both the personal norms and the psychology literature."
Interpersonal trust is primarily a private phenomenon that has caught the attention of social scientists and lawyers alike.12 Most trust scholars
7 Cf McAdams, supra note 4, at 360 (discussing exit from groups by members who disagree with a particular normative consensus).
8 See id. ("[G]roup discussion may produce ... a consensus.").
9 See, e.g., Michael P. Vandenbergh, Beyond Elegance: A Testable Typology of Social Norms in Corporate Environmental Compliance, 22 STAN. ENvrL. L.J. 55, 68-69 (2003) (out- lining the characteristics of internal norms).
10 See McAdams, supra note 4, at 349 (discussing how law can influence norms).
I See generally Erin Ann O'Hara & Claire Hill, Optimal Trust (unpublished manu- scriplt, on file with author).
See generally FRANCIS FUKUYAMA, TRUST: THE SOCIAL VIRTUES AND THE CREATION OF PROSPERITY (1995); TRUST: MAKING AND BREAKING COOPERATIVE RELATIONSHIPS (Diego Gambetta ed., 1988); TRUST & RECIPROcITY: INTERDISCIPLINARY LESSONS FROM EXPERIMENTAL RESEARCH (Elinor Ostrom & James Walker eds., 2003); Joyce Berg et al., Trust, Reciprocity, and Social History, 10 GAMES & ECON. BEHAV. 122 (1995); Margaret M. Blair & Lynn A. Stout, Trust, Trustworthiness, and the Behavioral Foundations of Corpo- rate Law, 149 U. PA. L. REV. 1735 (2001); Mark A. Hall, Law, Medicine, and Trust, 55 STAN. L. REv. 463 (2002); Stephen Knack & Philip Keefer, Does Social Capital Have an Economic Payoff? A Cross-Country Investigation, 112 Q.J. ECON. 1251 (1997); Rafael La Porta et al., Trust in Large Organizations, AM. ECON. REV., May 1997, at 333; Christopher R. Leslie, Trust, Distrust, and Antitrust, 82 TEX. L. REV. 515 (2004); Special Topic Forum on Trust in and Between Organizations, 23 ACAD. MGMT. REV. 387 (1998); PaulJ. Zak & Ste- phen Knack, Trust and Growth, 111 ECON. J. 295 (2001); JOHN W. DICKHAUT ET AL.,
1887
1888 UNIVERSITY OFPENNSYLVANIA LAWREVIEW [Vol. 153:1883
seem to agree that interpersonal trust is key to the healthy functioning of our social, political, and economic systems. While trust scholars disagree to some extent about the nature of trust, all seem to accept that trust involves a willingness to make oneself vulnerable to another despite a risk that the other will exploit that vulnerability. From a psychological perspective, trust as a concept fits comfortably into "the family of such notions as knowledge, belief, and the kind of judgment that might be called assessment."14 In short, trust is beginning to be understood as an essentially cognitive phenomenon.5
Once trust is conceptualized as a cognitive phenomenon, it be- comes possible to analyze trust as a judgment or assessment that is subject to the same types of biases to which other reasoning processes are subject. For example, the experimental literature indicates that people act as though they trust one another even in one-shot, anony- mous, experimental settings where lack of trust might instead seem appropriate. If these subjects can be said to "overtrust" from the per- spective of the rational actor, then we need an explanation for this sys- tematic tendency to behave irrationally. Evolutionary theorists posit that initial tendencies to cooperate, at least where the stakes are small, help to foster the development of cooperative relationships, which are essential to human survival. 16 Norms of cooperation, whether social or internal, can help to strengthen these tendencies. Other cognitive bi- ases appear to be present in the context of trust. For example, dis- trust of another person in one context can spill over into distrust of that person in other contexts as well as into distrust of others associ-17
ated with that person.'s When trust is treated as a cognitive phe-
TRUST, RECIPROCITY, AND INTERPERSONAL HISTORY: FOOL ME ONCE, SHAME ON YOU, FOOL ME TWICE, SHAME ON ME (Econ. Sci. Lab., Univ. of Ariz., working paper).
13 See Denise M. Rosseau et al., Not So Different After All: A Cross-Discipline View of Trust, 23 ACAD. MGMT. REv. 393, 395 (1998) (noting the widely held scholarly defini- tion of trust as "a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another" (emphasis omitted)).
14RUSSELL HARDIN, TRUST AND TRUSTWORTHINESS 7 (2002).
15 For a further explanation of this topic, see O'Hara & Hill, supra note 11.
16 See generally Robert Kurzban, Biological Foundations of Reciprocity, in TRUST &
RECIPROCITY, supra note 12, at 105.
17 See O'Hara & Hill, supra note 11 (manuscript at 4).
18 For a discussion of the difficulties associated with trusting members of an "out- group," see Roderick M. Kramer, Trust and Distrust in Organizations: Emerging Perspec- tives, Enduring Questions, 50 ANN. REv. PSYCHOL. 569, 588 (1999). On ingroup and outgroup attitudes generally, see Marilynn B. Brewer & Roderick M. Kramer, The Psy- chology ofIntergroup Attitudes and Behavior, 36 ANN. REV. PSYCHOL. 219 (1985).
2005] CHOICE OF LA WFOR INTERNET TRANSA CTIONS
nomenon, it also becomes possible to explore why trust is often slow to develop yet quick to break down.9 Similarly, the cognitive perspec- tive also enables us to predict and test the extent to which the average resilience of trust differs significantly across contexts.20 Some of these trust biases might turn out to serve important functions and therefore be "rational" in some larger or longer-term sense, but the point here is simply that we cannot always count on individuals to trust one an- other at a socially optimal level.
The policy implications of using law to foster trust appear more complex when the psychological perspective on trust is adopted. If trust were indistinguishable from cooperation and trade, then, in all contexts outside of illegal conspiracies,' the law should work to maximize trust in relationships. After all, cooperation and trade gen- erally appear to be unmitigated goods. Once we view trust as a cogni- tive assessment, however, it becomes clear that people are sometimes prone to trust too much and sometimes too little, depending on the social context of the trust assessment. Sweeping generalizations about the normative relationship between law and trust are no longer possi- ble; instead, more trust is desirable in some cases and less trust prefer- able in others. In short, legal scholars need to begin to think about how to promote "optimal trust,"22 and they must think hard about how trust is formed, maintained, and eroded in different types of relation- ships.
B. Trust and Law as Complements or Substitutes?
The positive relationship between law and trust is a subject of cur- rent commentary because scholars apparently disagree about the ex- tent to which law and trust are substitutes for or complements to one another. Their disagreement stems in part from the fact that scholars disagree over the applicability of trust in different interpersonal set- tings. Oliver Williamson, for example, has taken the position that trust is reserved only for interactions in close, personal relationships
19 See Blair & Stout, supra note 12, at 1776 (discussing the potential fragility of trust); Diego Gambetta, Can We Trust Trust?, in TRUST: MAKING AND BREAKING COOPERATIVE RELATIONS, supra note 12, at 213, 233-34 (discussing the fragility of trust but also the immunity of distrust to contrary experience).
20 See O'Hara & Hill, supra note 1 (manuscript at 5-6).
21 See Neal Kumar Katyal, Conspiracy Theory, 112 YALE L.J. 1307, 1325 (2003) (dis- cussing the benefits of trust in criminal conspiracies); Leslie, supra note 12, at 547
("Criminal conspiracies require trust just as beneficial cooperative ventures do.").
22 See generally O'Hara & Hill, supra note 11.
1889
1890 UNIVERSIFY OF PENNSYLVANIA LAWREVIEW [Vol. 153:1883
and has nothing to do with interactions in business contexts.2 3 Wil- liamson reaches this conclusion because he believes trust is relevant only in situations where people are interacting for noninstrumental reasons. When parties behave instrumentally, policymakers must in- stead focus their analysis on building institutional constraints (includ- ing law) on parties' opportunism.24 To Williamson, then, trust and law are not only substitutes for each other, they also cannot coexist as mo- tivational factors.
To other scholars, however, trust is present even in those contexts where behavior is instrumental.25 People often strive to structure their transactions and business relationships to maximize the likelihood that the parties behave appropriately, but despite these efforts, some duty is inevitably left unspecified, monitoring is expensive, and en- forcement is imperfect.26 In fact, to the extent that parties trust one another in a commercial context, they can conserve the transaction costs associated with structural constraints. 27 A broader conception of trust suggests that trust operates whenever the structural constraints cannot guarantee cooperative behavior. In these cases, the trust asses- sor must confront the possibility that the trust target might act in an
23 Oliver E. Williamson, Calculativeness, Trust, and Economic Organization, 36 J.L. &
ECON. 453, 469 (1993) ("I maintain that trust is irrelevant to commercial exchange and that reference to trust in this connection promotes confusion.").
24 Seeid. at 485 ("'[C]alculated trust' [is] a contradiction in terms ... .
25 Several scholars have taken Williamson to task for his position. See, e.g., Blair &
Stout, supra note 12, at 1748-50; Lawrence E. Mitchell, The Importance of Being Trusted, 81 B.U. L. REv. 591, 603-08 (2001); Lawrence E. Mitchell, Trust and Team Production in Post-Capitalist Society, 24J. CORP. L. 869, 888-91 (1999); Larry E. Ribstein, Law v. Trust, 81 B.U. L. REv. 553, 563-64 (2001). Moreover, Robert Frank has recently explored the possibility that nonconscious moral values such as trustworthiness serve important in- strumental ends, calling into question the power of the instrumental/noninstrumental dichotomy. ROBERT H. FRANK, PASSIONS WITHIN REASON: THE STRATEGIc ROLE OF THE EMOTIONS (1988).
26 See generally Ernst Fehr &John A. List, The Hidden Costs and Returns of Incentives- Trust and Trustworthiness Among CEOs, 2J. EUR. ECON. ASS'N 743 (2004).
27 Margaret Blair and Lynn Stout have written:
Where trust can be harnessed, it can substantially reduce the inefficiencies as- sociated with both agency and team production relationships. Trust permits transactions to go forward on the basis of a handshake rather than a complex formal contract; it reduces the need to expend resources on constant moni- toring of employees and business partners; and it avoids the uncertainty and expense associated with trying to enforce formal and informal agreements in the courts. Trust behavior also reduces losses from others' undetectable or unpunishable opportunistic behavior, losses that could discourage the forma- tion of valuable agency and team production relationships in the first place.
Blair & Stout, supra note 12, at 1757.
2005] CHOICE OF LAWFOR INTERNET TRANSACTIONS
untrustworthy fashion. When these gaps exist, and they almost always do, trust plays a role in parties' interactions. As Russell Hardin writes:
Giving people very strong incentives seems to move them toward being deterministic actors with respect to the matters at stake. At the other ex- treme, leaving them with no imputable reasons for action generally makes it impossible to trust them. Trust and trustworthiness (and choice and rationality) are at issue just because we are in the murky in-between.... 28
land that is neither deterministic nor fully indeterminate.
Note that from this alternative perspective the law and trust can work together in complementary fashion. Put differently, a legal rule can serve the function of giving a person an imputable reason for an- other's action. So long as the effect of the law is not so overpowering as to make the actor's choice deterministic, trust judgments work in the space between the two extremes described by Hardin to inform the trust assessor's willingness to interact with the target.
Norms also help to provide an imputable reason for a particular action.'9 Social norms are tied to rewards and sanctions that provide their own incentives for behavior. In contrast, internal norms of trustworthiness constrain an actor's willingness to behave opportunis- tically, even in the face of significant monetary or other temptations.30 Both social and internal norms (or their absence) thus work together to provide an "imputable reason" why it might be reasonable to pre- dict that an actor will act in a trustworthy (or an untrustworthy) fash- ion. The difference between the two types of norms, however, is that social norms work like law in that the reason a party abides by them is instrumental. As in the case of law, a strong enough social norm in- fluence could, at least in theory, obviate the trust inquiry. Internal norms of trustworthiness (along with assessments of competence), by contrast, are often the very focus of the trust assessment. Put differ- ently, to the extent that external constraints do not guarantee trust- worthy behavior, the assessor must ultimately determine the degree to which the actor is driven by internal norms of trustworthiness. A trustworthy person will be inclined to resist behaving according to her own short-term interests and instead will feel a commitment to do the
28 HARDIN, supra note 14, at 12.
29 See Jack Knight, Social Norms and the Rule of Law: Fostering Trust in a Socially Di- verse Society, in TRUST IN SOCIETY 354, 359 (Karen S. Cook ed., 2001) ("When the con- tent of the norms dictates cooperative behavior, social actors can use this information to develop expectations about the likelihood that others will cooperate .... ).
30 See FRANK, supra note 25, at 18, 69 (discussing the long-term benefits of this trustworthy behavior, which requires the short-term sacrifice of potential benefits).
1891
1892 UNIVERSITYOFPENNSYLVANIA LAWREVIEW [Vol. 153:1883 right thing, including to keep her promises. Of course, the tempta- tion to behave in an untrustworthy fashion can at some point become so great that the relatively trustworthy person submits to the tempta- tion." But the more trustworthy a person, the greater that temptation must be before the person will deviate from the path of trustworthy behavior.
The fact that law can influence social norms and that the two (law and social norms) together can have powerful influences on individu- als who lack internal norms of trustworthiness leads some to question whether law can in fact interfere with the development of trust. For example, Larry Ribstein recently took the position that law always in- terferes with trust because it has the effect of taking away the very vul- nerability that trust requires.2 From his perspective, the idea that law could ever work to enhance or foster interpersonal trust seems oxy- moronic.33
I propose a different conception of the relationship between law and trust. Let us label this conception the "safety net assessment."
Once we view trust as a cognitive assessment of one's willingness to make oneself vulnerable to another, then it becomes possible to think of the vulnerability in quantitative terms. People differ significantly in their willingness to trust others based on early life experiences.34 In addition, people trust some more than others, and they trust the same
31 SeeJay B. Barney & Mark H. Hansen, Trustworthiness as a Source of Competitive Ad- vantage, 15 STRATEGIC MGMT. J. 175, 179 n.3 (1994) ("In principle, some level of com- pensation will always exist where strong form trustworthy exchange partners will aban- don their values, principles, and standards of behavior, and act in opportunistic ways.").
32 See Ribstein, supra note 25, at 580-81 ("[Riegulation decreases the sense of vul- nerability that is critical to personal trust.").
33 See id. at 562 ("Law may dispose one party to rely on another because the other is subject to legal constraints. But this has nothing to do with the distinct concept of
trust.").
34 See generallyJulian B. Rotter, A New Scale for the Measurement of Interpersonal Trust, 35 J. PERSONALITY 651 (1967); Julian B. Rotter, Generalized Expectancies for Interpersonal
Trust, 26 AM. PSYCHOLOGIST 443 (1971); Julian B. Rotter, Interpersonal Trust, Trustwor- thiness and Gullibility, 35 AM. PSYCHOLOGIST 1 (1980). Those who grew up in privi- leged, sheltered environments tend to enter the adult world with high levels of trust in others. In contrast, those who were abused and/or severely neglected as children tend to enter the adult world with very low levels of trust in others. HARDIN, supra note 14, at 116-19. Because high-trust individuals are more willing to interact with others in meaningful ways than are low-trust individuals, those who place too much trust in oth- ers are more likely to receive feedback suggesting that their assessments need to be modified than those who place too little trust in others. Id. at 119-24.
2005] CHOICE OF LAW FOR INTERNET TRANSACTIONS
people differently in different contexts.35 Most, but not all, interper- sonal relationships start out with relatively little trust simply because
the parties lack trust-relevant information about each other. Typi- cally, the parties take small trust steps initially while only gradually in- creasing the degree to which they are inclined to make themselves vulnerable to one another.36 Moreover, even in relationships that ob-
tain high trust levels, the parties often also must contend with pockets of distrust.37 For example, I might trust my friend with my children and my car but distrust the likelihood that she will pay back money or return clothes that she has borrowed. Arenas of trust and distrust dif- fer from relationship to relationship, but the important point here is that both trust and distrust are informed by the experiences that peo- ple have with other people in general and the trust target in particu- lar.
With each trustjudgment, the evaluator must assess whether she is willing to make herself vulnerable by plunging off a metaphorical cliff.
Notice that the degree to which she is willing to make herself vulner- able turns on: (1) her own propensity to trust others; (2) the degree of interpersonal trust in the general environment or society;38 and (3) her assessment of the other's trustworthiness. Moreover, people seem to make decisions about whether to trust another based on the extent to which the other can be trusted to (1) not act against the assessor's interests, and (2) act with basic competence. If the assessor has com- plete trust in the other person, she is willing to take a step off the edge of the cliff regardless of the vertical distance between the edge and the ground. In her complete trust, she is confident that before she hits the ground a very thick, sturdy, and soft safety net will cushion her fall. When she trusts completely, she does not perceive a risk that she might be hurt as a consequence of her interaction, even though her vulnerability might be great. In contrast, if she affirmatively distrusts another person, she might feel relatively certain that if she takes a step off the cliff then she will fall a fair distance with no net at all to cush- ion her fall. At the beginning of a relationship, when the assessor lacks reliable trust information, the assessor must determine the ex-
35 See Roy J. Lewicki et aL., Trust and Distrust: New Relationships and Realities, 23 AcAD. MGMT. REv. 438, 442 (1998) ("Within the same relationship we have different encounters in different contexts with different intentions that lead to different out- comes.").
See O'Hara & Hill, supra note 11 (manuscript at 22-23, 32).
37 See Lewicki et al., supra note 35, at 443-44 (noting that parties often compart- mentalize their relationships).
38 See generally FUKUYAMA, supra note 12.
1893
1894 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 153:1883
tent to which she is willing to fall in the event that she mistakenly trusts. The assessor may be willing to take the risk of a fairly short drop-say a step or two down-but she is likely unwilling to make her- self vulnerable to falling much farther because she lacks information about the reliability of the safety net.
Under this safety-net conception, the external constraints-law and social norms-are serving to add fortification to the safety net. If legal protections are weak, monitoring difficult, and sanctions hard to impose, then very little is added to the strength of the safety net. If legal and/or social incentives are instead powerful constraints on be- havior, then they can provide so thick and sturdy a safety net that the assessor is no longer inclined to worry about the integrity of the ac- tor's net standing alone. In fact, if legal and social sanctions were per- fectly effective in influencing behavior, then the actor would have no incentive to invest in fortifying his own net-through reputation, goodwill, or otherwise. Instead, he knows that the assessor will rely solely on the availability of the net built by society.
The assessor must determine whether or not to undertake a verti- cal drop associated with an interaction. Her willingness to take the plunge turns both on the strength of the net (which is affected by law, social norms, and an assessment of the actor's trustworthiness) and the magnitude of the vertical drop associated with the plunge. If the vulnerability is perceived to be quite small, then the drop is shallow, and the assessor will often be willing to take the plunge regardless of the strength of the safety net. Put differently, small potential risks are often undertaken without any information about legal sanctions or the other's trustworthiness.40
To the extent that interactions are lumpy, however, so that the plunge cannot be confined to de minimis vertical falls, then the asses- sor may be unwilling to take the plunge due to lack of information about the investment of the actor in the strength of the net (his trust- worthiness). To the extent that law can work to help fortify the net, the assessor becomes more willing to take the plunge. Social sanc- tions can have the same effect, at least to the extent that the assessor and the actor can be said to belong to the same group. But in many cases, the plunge will not be taken without the safety net that law pro- vides. And by providing an incentive for the assessor to take the first
39 Cf Ribstein, supra note 25, at 555 (arguing that mandatory laws can crowd out trust).
Presumably this is the cooperation among strangers observed in many social di- lemma game experiments.
2005] CHOICE OF LA WFOR INTERNET TRANSACTIONS
plunge, the law can make it possible for the creation of interpersonal trust in that relationship-i.e., discovery that the other party's net is quite safe-that is strong enough to make the parties take greater plunges in the future.41
C. The Truster's Copayment?
The challenge for policymakers, it seems, is to find a way for the law to help fortify the safety net available to the assessor without re- moving the incentive of the actor to invest in his own safety-net fortifi- cation. That is, the law must aim to encourage the development of in- ternal norms of trustworthiness while at the same time encouraging assessors who lack trust-relevant information to nevertheless take the plunge. The optimal trust-fortification regime might therefore be one under which the government acts as imperfect insurer of trust deci- sions with legal rules and sanctions, but the assessor is effectively forced into a copay arrangement under which she must incur some,42
but not all, of the costs of mistakenly trusting. Ideally the copay amount is small enough not to discourage initial interactions, yet large enough so that the assessor is vigilant to act on cues-i.e., repu- tation-which suggest that distrust is more appropriate. By encourag- ing continued vigilance by the assessor, the copay arrangement causes the actor to continue to invest in trustworthiness.
Notice that the optimal amount of the trust assessor's copayment turns critically on several factors, including but certainly not limited to the relationship of the parties (which reflects the extent to which their preferences are interdependent), the degree to which the assessor and the actor can be said to belong to the same social group (where social norms also constrain behavior), the ability of the assessor to otherwise obtain information about the actor's reputation for trust- worthiness, the likelihood that the actor must make parallel trust as- sessments about the assessor, and the overall degree of vulnerability that the interaction represents for the parties.43
In contrast to arm's-length transactions, fiduciary relationships that are one-sided in trust can cause greater difficulties for the parties
41 Ribstein dismisses without explanation this role of law in promoting trust. Rib- stein, supra note 25, at 564.
42 1 am indebted to Claire Hill for pointing out that the optimal structure here is
identical to the medical copay arrangement for patients.
45 If the relationship is of a type that is characterized as one of mutual high or thick trust, as is the case with family, then the law need not step in as insurer at all.
1895
1896 UN!VERSFTY OF PENNSYLVANIA LAW REVIEW [Vol. 153:1883
entering them. For example, once a doctor-patient or a lawyer-client relationship is formed, the client or patient very often places signifi- cant trust in the competence and loyalty of the lawyer or doctor. If, however, the patient or the client knows that others have been signifi- cantly harmed by trusting the doctor or the lawyer, then the trust nec- essary to the development of these relationships is eroded. In this context, the law places special duties of loyalty and care on the trusted person to signify the importance of his trustworthiness. Moreover, both the medical and legal fields are regulated and monitored by or- ganizations charged with ensuring the delivery of high-quality services.
In these cases, the aim of the law presumably is to keep the copayment low enough so that people feel comfortable entering these relation- ships of vulnerability.
The challenge, as always, is to avoid making the fiduciary duties so rigorous that professionals are afraid to vigorously advocate for their patients' and clients' well-being. If the duties are too onerous, the professionals practice defensively rather than optimally. On the other hand, once it is determined that the professional has violated his duty, the law should work hard to fully compensate the patient or client for the harm that she suffered as a result of the breach. In any event, ce- teris paibus, we would tend to expect that the copayment required for each party in a legal relationship will vary inversely with the vertical height of the plunge and the ability (and desirability) of the parties to gather and disseminate their own information about the other party's trustworthiness.
In summary, so long as legal and social sanctions working together are not so powerful that they guarantee full contract performance by each party, then law and trust can work together in complementary fashion. To some extent, however, law and trust function as substi- tutes in that legal protections can work to decrease (if not eliminate) the extent of vulnerability associated with contracting. To the extent that the law works to decrease rather than eliminate contracting par- ties' vulnerability, or copayment, legal rules can work to promote in- terpersonal trust by increasing parties' willingness to initiate a smaller- scale contractual relationship where they would not otherwise be will- ing to contract. To the extent that these contracting experiments go well, the parties are more willing to embark on longer-term and higher-value contractual relationships. In these circumstances, law and trust begin as substitutes, but overall we might expect the two to end up working in complementary fashion.
2005] CHOICE OF LA W FOR INTERNET TRANSACTIONS II. ONLINE TRUST BY CONSUMER PURCHASERS
This Part argues that consumer trust is essential for online trans- actions, that consumer trust in some ways is more difficult to obtain by online vendors than by brick-and-mortar vendors, and that unknown vendors are likely to suffer most from problems of consumer trust. If consumers are unwilling to experiment with unknown vendors, then the Internet fails to create competitive pressures that serve to erode online price dispersions between vendors. When competitive pres- sures are relaxed, consumers pay more for their products, and con- sumer welfare is thus not maximized. This Part further argues that applicable contract law can help to foster consumer trust in online vendors, but choice-of-law decisions need to help facilitate the incor- poration of trust-enhancing rules for online consumer transactions.
Parts III and IV, respectively, will explore the content of trust- enhancing contract rules and the difficulties of obtaining them on an international level.
A. Trust and Contract
To what extent should contract law be shaped by trust concerns?
In distinguishing fiduciary from mere contract relationships, com- mentators often focus on the fact that fiduciary relationships are trust relationships, whereas contract relationships typically are of the arm's- length, caveat emptor type." From this perspective, one might con- clude that contract law itself has nothing to do with trust. I take a dif- ferent position here, by asserting that fiduciary obligations are just one of the many situations where lawmakers worry about interper- sonal trust in relationships.
Duties of care and loyalty are imposed on professionals who enter relationships as fiduciaries. These duties are designed to encourage trust on the part of the nonprofessional while simultaneously encour- aging trustworthy behavior on the part of the professional. Fiduciary relationships often cannot work effectively unless the nonprofes- sional's trust assessments are fundamentally noncalculative. Put dif- ferently, in fiduciary relationships one of the parties often must place herself in a position of blind trust regarding the loyalty and compe- tence of the professional. Without unquestioning trust, the nonpro- fessional might fail to disclose information that is valuable to the con-
44 See, e.g., Bain v. Champlin Petroleum Co., 692 F.2d 43, 47 (8th Cir. 1982);
Meinhard v. Salmon, 164 N.E. 545, 546 (N.Y. 1928).
1897
1898 UN1VERSTFY OFPENNSYLVANIA LAWPREVIEW [Vol. 153: 1883
ferral of high quality professional services.5 Moreover, at least in the medical field there is some evidence that blind trust works as a pla- cebo to improve the patient's condition even before the doctor ren- ders her aid4 6 If patients and clients are worse off in environments where they consciously process trust-relevant information about the professional during the course of the professional's service provision, then fiduciary duties become one of the law's methods of promoting an environment where blind, or noncalculative, trust follows.47
In contrast, nonfiduciary contractual relationships are, by defini- tion, not relationships that policymakers believe need to be fostered by blind or unquestioning trust. Instead, arm's-length relationships are those where the parties are encouraged to scrutinize each other and pay attention to and bargain for those terms that each believes are important to the transaction. In this setting, the law promotes conscious calculations about both the trustworthiness of the other and the duties that each is undertaking. One of the ways that the law promotes conscious calculation is by ensuring that trust assessors are left with a copayment in the event the other party behaves in an un- trustworthy fashion. In this sense, one might conclude that contract law does not serve to promote trust.
Nonetheless, contract law in general serves the important safety- net function described in Part I in that it enables strangers to be will- ing to interact, potentially to their mutual advantage. Without a reli- able contract law system, parties are inclined to do business only with family, friends, and close community members.4 Much of the value of
45 Cf Miriam J. Metzger, Privacy, Trust, and Disclosure: Exploring Barriers to Electronic Commerce, J. COMPUTER-MEDIATED CoMM., July 2004, at http://jcmc.indiana.edu/
vol9/issue4/metzger.html ("Trust is perhaps the most important influence on infor- mation disclosure .... [It] is a precondition for self-disclosure because it reduces the perceived risks involved in revealing private information.... .").
46 See Mark A. Hall, Law, Medicine, and Trust, 55 STAN. L. REV. 463, 479 (2002) ("[T]he doctor himself is a placebo .... "); Erin Ann O'Hara, Apology and Thick Trust:
What Spouse Abusers and Negligent Doctors Might Have in Common, 79 CHI.-KENTT L. REv.
1055, 1080 (2004) ("[Tlrust in the physician can have a powerful placebo effect on patients.").
47 Presumably, physician licensing requirements and medical malpractice liability are the other legal methods for encouraging high levels of patient trust. There are several nonlegal mechanisms that also work to promote patient trust in their doctors.
See O'Hara, supra note 46, at 1081 (noting that doctors' practice behaviors tend to em- phasize both the doctors' elevated social status and competence and the patients' vul- nerability, which tend to increase patient trust).
48 SeeJanet T. Landa, A Theory of the Ethnically Homogeneous Middleman Group: An Institutional Alternative to Contract Law, 10 J. LEGAL STUD. 349, 351 (1981) ("Under conditions of contract uncertainty... a rational trader will not indiscriminately enter
2005] CHOICE OF LA W FOR INTERNET TRANSACTIONS
a contractual relationship lies in the fact that it enables each of the parties to rely on the contract and to therefore move on to address other concerns for the future. Without contract law, each party must very carefully assess the extent to which the other is reliable, and many possibly advantageous trades therefore would not occur without sig- nificant information about the trustworthiness (and reliability) of the other person. An effective contract law system serves the function of providing a safety net that is sufficiently strong so that it makes people willing to contract with acquaintances and strangers-to take the leap of faith-at least for small-value transactions. By expanding the possi- ble trading partners for each person, contract law makes us all wealth- ier because it provides greater possible gains from exchange.
In addition, contract law to some extent serves the egalitarian function of providing incentives for each person to branch out of her tight-knit, cliquish community to deal with people of different classes, backgrounds, and beliefs. Of course, the law cannot go too far in this direction if we simultaneously wish to satisfy other goals. In particular, the ultimate goal of contract law must be to help fortify, rather than replace, the safety net of the parties. Contract law provides incentives for strangers to be willing to contract with one another, but ultimately we are all better off if people still have incentives to be trustworthy- that is, to be reliable and hardworking. To keep this incentive in place, the law cannot guarantee full compensation for breach of promises, even where we believe that the breacher had a moral obliga- tion to keep the promise.
Consider, for example, a legal system that guaranteed full com- pensation in the event of a breach of promise. To be effective, that system would have to award plaintiffs enough to compensate them for their attorney and court fees, and for the opportunity costs of their time spent mitigating damages and seeking relief. Moreover, that sys- tem would have to insure the promise in the event that the promisor
into impersonal exchange relations with anonymous traders."); Janet T. Landa & Xiao Tian Wang, Bounded Rationality of Economic Man: Decision Making Under Ecological, Social, and Institutional Constraints, 3J. BIOECON. 217, 227 (2001) (noting that Chinese rubber merchants-who operate under an uncertain contract scheme-tend to choose as trading partners "kinsmen, friends, people from the same place in China and those who speak the same dialect").
49 Even in Europe, where specific performance is nominally the primary remedy, parties routinely seek and receive money damages for breach of contract. See generally HENRIK LANDO & CASPAR ROSE, THE MYrH OF SPECIFIC PERFORMANCE IN CIVIL LAW COUNTRIEs (Am. Law & Econ. Ass'n Annual Meetings, Paper No. 15, 2004), available at http://law.bepress.com/alea/14th/artl5/.
1899
1900 UNIVERSITY OFPENNSYLVANIA LAWREVIEW [Vol. 153:1883
cannot be made to answer in damages for the breach by reason of bankruptcy, lack of funds, inability to attach assets, or inability to lo- cate the promisor. A strong-form guarantee of compensation might well prove expensive for the government, especially as people slowly become indifferent to the trustworthiness of those with whom they contract. The insurance function of the government would increase over time under that system.
With an imperfect safety net, the promisee still has an incentive to pay attention to trust-relevant information about the promisor, and by paying attention to that information, the promisors of the world re- tain an incentive to perform as they have promised. In the process, the imperfect safety net might also have the effect of biasing people's choices of contract partners in favor of those about whom they believe they have reliable trust-relevant information. We ask our friends, neighbors, colleagues, and churchmates for advice about where to get needed goods and services, and employers often consult trusted re- commenders before making hiring decisions. In short, a well- functioning contract law system can encourage significant trade with strangers but will likely fail by itself to produce radical shifts toward egalitarianism.
The adoption of trust-enhancing contract rules, even where po- tentially valuable, must proceed cautiously, for several reasons. First, to craft doctrines that discourage all untrustworthy (including unreli- able) behavior, contract law must abandon formal rules and move to- ward standards of good faith, reasonableness, and fairness.50 Other- wise, the safety net would not take the shape necessary to provide optimal incentives. On the other hand, it is generally understood that vague standards create costly uncertainties for contracting parties.5' The imposition of vague standards might be avoidable in those con- tract settings where the parties have detailed knowledge of one an-
50 For example, the consideration rule for contract modification has been re- placed by inquiries into "good faith" and "legitimate commercial reason" under the Uniform Commercial Code, U.C.C. § 2-209 cmt. 2 (2004), and into an inquiry about what is "fair and equitable" and about what 'justice requires" under the RESTATEMENT
(SECOND) OF CONTRACTS § 89 (1981). Similarly, exclusive dealing requirements now require "best efforts" under the Uniform Commercial Code. U.C.C. § 2-306(2).
51 See, e.g., Douglas G. Baird & Robert Weisberg, Rules, Standards, and the Battle of the Forms: A Reassessment of§ 2-207, 68 VA. L. REv. 1217, 1227-28 (1982) (noting that standards, as opposed to rules, might "breed uncertainty" and "discourage people from entering into transactions"); Frank H. Easterbrook, Afterword: On Being a Commer- cial Court, 65 CHI.-KENT L. REv. 877, 877 (1989) ("A good commercial judge prefers rules to standards, for even when a rule is flawed people may form contracts against a known background.").
2005] CHOICE OF LAW FOR INTERNET TRANSA CTIONS
other's reputations and the ability to use social sanctions to induce compliance with appropriate standards of behavior.52 In any event, the shift from traditional common law contract rules toward stan- dards5 no doubt reflects the fact that over time more contracting par- ties found themselves interacting with strangers on the basis of limited information and that the law sought to validate and encourage these relationships. 54 The challenge for courts and legislatures is to use con- tract law to enhance trust while minimizing where possible its resort to vague standards.
Second, legal reformers must be careful that they not deliver to the protected party a sword along with its shield. Put differently, to the extent that contracting parties cannot know ex ante which one of them is at greater risk of breach, the law must be careful in protecting one party that it not empower that party to use its protection in an un- trustworthy fashion.5 Consider, for example, the parol evidence rule.
A court might be concerned that the parol evidence rule enables a party-call him A-to make oral representations that cause reliance by. the other party-call her B-but that ultimately are unenforceable.
To enhance the willingness of B to enter a contract, the court could relax the parol evidence rule and enable oral evidence to be admitted in court. Once oral evidence is admissible at trial, however, the pro-
52 Some commercial parties are effectively able to ignore or avoid governing con- tract law through private sanctions and/or private dispute resolution. See, e.g., Lisa Bernstein, Opting Out of the Legal System: Extralegal Contractual Regulations in the Diamond Industry, 21 J. LEGAL STUD. 115, 115 (1992) (noting that diamond traders "have devel- oped an elaborate, internal set of rules, complete with distinctive institutions and sanc- tions, to handle disputes among industry members"). The more general point empha- sized in the text is that rules might be appropriate for some contractual settings even though standards must be resorted to in others.
5' See, e.g., E. Allan Farnsworth, Some Prefatory Remarks: From Rules to Standards, 67 CORNELL L. REv. 634, 634 (1982) (examining the bases for this shift); Richard E. Spei- del, Afterword: The Shifting Domain of Contract, 90 Nw. U. L. REv. 254, 260-61 (1995)
(documenting the existence of this shift); see also Mark P. Gergen, The Jury's Role in De- ciding Normative Issues in the American Common Law, 68 FORDHAM L. REV. 407, 442 (1999) (questioning whether the shift from rules to standards in contract law will in- crease the rble of the jury).
54 Cf Lynne G. Zucker, Production of Trust: Institutional Sources of Economic Structure, 1840-1920, 8 RES. ORGANIZATIONAL BEHAv. 53, 53 (1986) (suggesting that high rates of immigration, internal migration, and business enterprise instability during the late nineteenth and early twentieth centuries led to a disruption of informal interpersonal trust and to the necessity of providing alternative institutional-based trust measures, including government regulation and legislation regarding transactions).
55 See Ribstein, supra note 25, at 576 ("[I] mposing extra duties to reduce the par- ties' vulnerability to the risk of disappointment may increase their vulnerability to op- portunistic litigation.").
1901
